Jochelson (Reference Jochelson2006) highlights the very important challenges that mental health units in the UK are likely to face in becoming smoke-free environments. Although there is very little doubt about the benefits of protecting patients and staff from the direct and indirect effects of smoking, the crude application of regulations of the English Health Act 2006 to all psychiatric settings might not be entirely beneficial and some patients might need to be exempt. Individuals presenting with severe psychopathology, those lacking capacity to agree to nicotine replacement treatment and individuals admitted under the Mental Health Act 1983 who have reduced civil liberties and limited access to outdoor space raise considerable concerns. Under these circumstances a forced nicotine withdrawal is likely. This iatrogenic phenomenon is associated with significant risks such as severe exacerbation or misinterpretation of psychiatric symptoms (Reference Greeman and McClellanGreeman & McClellan, 1991; Reference Dalak and Meador-WoodruffDalak & Meador-Woodruff, 1996), and pharmacokinetic changes resulting in increased concentration of psychotropic medications (Reference HughesHughes, 1993).

Jochelson minimises concern that under these circumstances there might be an increased risk of aggressive behaviour in psychiatric patients. The reality is that it is very difficult to be certain because the literature offers controversial findings. In older studies, which report negative results, the information is mostly retrospective and qualitative, and studies have adopted different outcome measures and failed to control for a number of fundamental variables such as access to the outside, which may vary according to staff availability and patient status (e.g. under the Mental Health Act 1983), hospital setting (in-patients, out-patients, intensive care units, etc.), psychiatric diagnosis, degree of psychopathology, level of dependence, comorbidity with other addictive behaviours, motivation, etc. (For review see Reference El-Guebaly, Cathcart and CurrieEl-Guebaly et al, 2002.) This has resulted in the limited generalisability of the findings. More recent studies have controlled for these variables and have reported increased irritability and agitation among psychiatric patients, with disengagement from services and premature discharge (e.g. Reference Prochaska, Gill and HallProchaska et al, 2004). It is also noteworthy, if the ban is intended to enhance the long-term health of psychiatric patients, that experience emerging from other countries where smoking bans in psychiatric hospitals have already been implemented suggests that resumption of smoking after discharge is the most likely outcome, with questionable long-term effects (Reference El-Guebaly, Cathcart and CurrieEl-Guebaly et al, 2002; Reference Lawn and PolsLawn & Pols, 2005; Reference Prochaska, Fletcher and HallProchaska et al, 2006).

Effective measures to increase the chance of positive health benefits could be based on evidence emerging from the treatment of nicotine addiction in hospitalised patients. An effective strategy includes diagnosis and treatment planning with nicotine replacement therapy or bupropion, on-unit dedicated smoking cessation counselling, reasonably extensive behavioural support, and post-discharge referral for treatment of nicotine dependence (Reference WestWest, 2002). Eliminating the burden of tobacco use in psychiatric hospitals is a public health priority but must be delivered in such a way that risks are minimised in otherwise vulnerable individuals and healthcare systems are developed that are capable of delivering effective treatments.